Endocarditis due to vancomycin-resistant Enterococcus faecium in an immunocompromised patient: cure by administering combination therapy with quinupristin/dalfopristin and high-dose ampicillin.

Abstract: A 56-year-old man with diabetes mellitus and cadaveric renal transplantation had vancomycin-resistant Enterococcus faecium tricuspid valve endocarditis. Relapse followed 6 weeks of treatment with intravenous gentamicin and high-dose ampicillin. On the basis of previous data suggesting the potential for synergistic activity of quinupristin/dalfopristin plus high-dose ampicillin, therapy with this combination was administered for 63 days. Cure was achieved and later confirmed at 2-year follow-up.

**********

Key Points

* Vancomycin resistance in Enterococcus faecium has increased the challenge presented by enterococcal endocarditis, because prolonged, bactericidal therapy is necessary for cure.

* Although quinupristin/dalfopristin, an injectable streptogramin antibiotic, generally exerts bacteriostatic activity against vancomycin resistance in Enterococcus faecium, a combination of quinupristin/dalfopristin with a [beta]-lactam antibiotic such as ampicillin can be synergistic against some strains.

* After failure of a 6-week course of gentamicin and high-dose ampicillin, a prolonged course of quinupristin/dalfopristin plus high-dose ampicillin cured a case of vancomycin-resistant Enterococcus faecium tricuspid valve endocarditis in a diabetic man with a cadaveric renal transplant.

Case Report

A 56-year-old man with diabetes mellitus type 1, chronic renal insufficiency, and cadaveric renal transplantation had a 6-week history of sweats and anorexia, followed by 2 weeks of chills and fever, with temperatures up to 38.4[degrees]C (101.1[degrees]F). Two blood cultures yielded Gram-positive cocci in chains. Vancomycin therapy (1 g administered intravenously daily) was initiated (Day 1). Relevant chronology of blood culture results and antimicrobial therapy are shown in Figure 1.

[FIGURE 1 OMITTED]

The pathogen was identified as E. faecium resistant to vancomycin, ampicillin-sulbactam, and ciprofloxacin; borderline susceptible to doxycycline; and susceptible to chloramphenicol by disk diffusion testing. Gentamicin was inhibitory at 120 mg/ml, suggesting possible synergy with a cell-wall active agent against enterococci at clinically achievable serum concentrations. Broth dilution testing results are shown in Table 1.

Physical examination revealed a temperature of 37.4[degrees]C (99.3[degrees]F), weight 121 kg, normal cardiac function, no embolic phenomena, and multiple absent teeth but no abscess. Medications included doxycycline, prednisone (8 mg/d), cyclosporine (100 mg bid), and a diuretic. A transesophageal echocardiogram showed an irregular vegetation, greater than 1 cm in maximal dimension, adherent to the nonseptal leaflet of the tricuspid valve.

Bacteremia persisted despite treatment with chloramphenicol and doxycycline; chloramphenicol peak concentration was 12.1 mg/ml. Ampicillin and gentamicin were substituted for doxycycline and chloramphenicol (Fig. 1), after which symptoms resolved. Potential synergy of ampicillin and gentamicin in combination was assessed using the checkerboard microtiter method. The fractional inhibitory concentration (FIC) for each drug (ampicillin or gentamicin) was calculated by dividing the concentration of that drug necessary to inhibit growth in a given row or column of the microtiter plates by the minimum inhibitory concentration of the test organism to that drug alone; the FIC index was obtained by adding the FIC value for ampicillin to that for gentamicin. The FIC index was interpreted as synergistic if [less than or equal to] 0.5, additive if 1, indifferent if 2 or 4, and antagonistic if greater than 4. In this case, the FIC index was 0.8, suggesting synergy. On Day 54, antibiotic therapy was discontinued.

Bacteremia recurred on Day 82, and treatment with ampicillin and gentamicin was restarted (Fig. 1). Repeated transesophageal echocardiogram showed an irregular mass, slightly larger than on previous studies. Serum was inhibitory at a titer of 1:16, but not bactericidal at 1:2. Because the vancomycin-resistant E.faecium was susceptible to Q/D (Table 1), Q/D therapy was started at 960 mg (7.5 mg/kg) intravenously every 8 hours (Fig. 1). Ampicillin therapy was maintained. Serum was inhibitory at a titer of 1:16, though not bactericidal at 1:2.

On Day 109, myalgias in the shoulder girdle and proximal upper extremities were partially controlled with oxycodone and diazepam. On Day 129, the patient was hospitalized because of hypotension, increasing creatinine level (4.9 mg/dl), and a serum sodium value of 130 mEq/L. Myalgias continued. The Q/D was withheld, and the ampicillin dose was reduced (Fig. 1). All findings improved rapidly. On Day 143, Q/D therapy was reinitiated at 600 mg (5 mg/kg) every 8 hours. Mild lower lumbosacral myalgias were moderately relieved by methocarbamol, oxycodone, and diazepam. A serum inhibitory titer was 1:16, with bactericidal activity at 1:4. Total bilirubin was slightly elevated (1.3 1.4 mg/dl). On Day 180, ampicillin and Q/D were discontinued (Fig. 1). The patient lost 10 lb despite lower diuretic dosing. The mild hyponatremia and elevated bilirubin resolved. Follow-up at 24 months confirmed cure.

The development of vancomycin-resistant Enterococcus faecium has increased the therapeutic challenge presented by enterococcal endocarditis, because prolonged bactericidal therapy is necessary for cure. We report a case of tricuspid valve endocarditis due to vancomycin-resistant E. faecium in an immunocompromised patient who was successfully treated with quinupristin/dalfopristin (Q/D) and high-dose ampicillin after unsuccessful gentamicin and high-dose ampicillin therapy.

Discussion

After the failure of a 6-week course of gentamicin and high-dose ampicillin, a prolonged course of Q/D with high-dose ampicillin cured this well-documented case of tricuspid valve endocarditis due to vancomycin-resistant E. faecium. Cure was obtained despite the persistence of a large vegetation. Although Q/D generally exerts bacteriostatic activity against vancomycin-resistant E. faecium, a combination of Q/D with a cell-wall active antibiotic such as ampicillin can be synergistic against some strains. (1) Given the lack of therapeutic alternatives and the inadvisability of surgical intervention, Q/D and high-dose ampicillin were administered to our patient. The patient's isolate exhibited an ampicillin minimum inhibitory concentration of 32 mg/ml, indicating the potential for achieving a therapeutic effect with high-dose therapy, and the Q/D minimal bactericidal concentration was relatively low (1.0 mg/ml); therefore, clinical efficacy might have been driven by Q/D alone, a synergistic combination with ampicillin, or simply an extremely prolonged treatment duration. It is also possible that prolonged retreatment with ampicillin and gentamicin might have resulted in cure.

Dosage adjustment of Q/D is not recommended in patients with obesity or renal insufficiency. (2) However, the dose of 960 mg every 8 hours initially used for this patient is at the upper end of that studied in Phase III trials (data on file, Aventis Pharmaceuticals, Inc.) and might have contributed to development of myalgias. (2) Use of analgesics and benzodiazepines, along with temporary interruption of Q/D and reinstitution thereof at a lower dose, allowed a full course of treatment. (2)

Because Q/D is administered in 100 to 500 ml dextrose (5%) in water, the occurrence of fluid overload and hyponatremia in a patient with chronic renal insufficiency could be anticipated. (2) Mild hyperbilirubinemia may have been caused by interference with bilirubin secretion. (2)

This case provides anecdotal confirmation of the ability of Q/D to penetrate vegetations (3) and augments a previous report of successful treatment of vancomycin-resistant E. faecium endocarditis with Q/D plus doxycycline and rifampin, also shown to be a synergistic drug combination. (4) A therapeutic failure of Q/D in a central venous thrombosis infected with vancomycin-resistant E. faecium could have been due to the use of chloramphenicol and doxycycline, but indicates that linezolid could be useful in some such cases. (5) In summary, prolonged treatment with Q/D and high-dose ampicillin cured E. faecium endocarditis in a severely immunocompromised patient with a large tricuspid valve vegetation.

Table 1. Broth dilution susceptibility testing (a)

Antibiotic                    MIC ([micro]g/ml)    MBC ([micro]g/ml)

Ampicillin                           32                   >64
Chloramphenicol                       8                   >64
Doxycycline                          16                    64
Imipenem                             32                   >64
Quinupristin/dalfopristin             0.5                   1

(a) MIC, minimum inhibitory concentration; MBC, minimum bactericidal
concentration: NA, not applicable.

References

(1.) Matsumura SO, Louie L, Louie M, Simor AE. Synergy testing of vancomycin-resistant Enteroeoccusfaecium against quinupristin-dalfopristin in combination with other antimicrobial agents. Antimicrob Agents Chemother 1999;43:2776-2779.

(2.) Rubinstein E, Prokocimer P, Talbot GH. Safety and tolerability of quinupristin/dalfopristin: Administration guidelines. J Antimicrob Chemother 1999;44:37-46.

(3.) Fantin B, Leclercq R, Ottaviani M, Vallois JM, Maziere B, Duval J, et al. In vivo activities and penetration of the two components of the streptogramin RP 59500 in cardiac vegetations of experimental endocarditis. Antimicrob Agents Chemother 1994;38:432-437.

(4.) Matsumura S, Simor AE. Treatment of endoearditis due to vancomycin-resistant Enterocoecus faecium with quinupristin/dalfopristin, doxycycline, and rifampin: A synergistic drug combination. Clin Infect Dis. 1998; 27:1554-1556.

(5.) McNeil SA, Clark NM, Chandrasekar PH, Kauffman CA. Successful treatment of vancomycin-resistant Enteroeoccusfaecium bacteremia with linezolid after failure of treatment with Synercid (quinupristin/dalfopristin). Clin Infect Dis 2000;30:403-404.

From the Infectious Disease Section, Group Health Cooperative of Puget Sound, Seattle, WA; and Aventis Pharmaceuticals, Inc., Bridgewater, NJ.

Quinupristin/dalfopristin was provided by Aventis Pharmaceuticals, Inc. (previously Rhone-Poulenc Rorer Pharmaceuticals, Inc.) to Dr. Thompson for the treatment of the patient presented in this report, and Drs. Talbot and Lavin were employees of Aventis when this patient was treated. Dr. Talbot is a paid consultant to Aventis.

Reprint requests to Robert L. Thompson, MD, Infectious Disease Section, Group Health Cooperative of Puget Sound, 125 16th Avenue E., CSB-2, Seattle, WA 98112. Email: [e-mail]

Accepted October 24, 2001.